Inverse solution techniques for determining aquifer parameters from pump test data

Abstract

The estimation of aquifer parameters is often difficult, inaccurate and time consuming. A method of analysing pumping test data is presented which automatically estimates aquifer parameters. The inverse solution technique of least squares is used to obtain the parameters which give the best fit between the observed and theoretical values of drawdown. Three different solution methods are presented. Data from a confined aquifer is analysed using the Theis solution. For leaky aquifer data, either the Walton or Hantush solution may be used. The least squares algorithms for the leaky condition were combined with an Hermitian interpolation algorithm to reduce the running time. In each case the algorithm converged from a range of values to the connect solution. The algorithms were validated by comparing the results of well documented pumping tests with the results of other authors. A programme of fieldwork was conducted which was analysed using the algorithms. The results of this analysis were further validated by numerical modelling. The pumping test was simulated using the calculated values of aquifer parameters. The theoretical drawdown curve matches very closely to the field values observed during the pumping tests.